[TCML] Current Along Secondary Length

[bartb]

Hi all,
(changed the subject line to better serve the subject)

I modeled the demo coil of Javatc with and without a top load plotting 
the current and voltage along the length of the coil) and graphed the 
data in Excel. Uses a 12/60 NST size with a 18.8nF cap.

The first case is a coil without a top load. Note how the current 
changes to almost nil at the top of the coil from the initial base current.
http://www.classictesla.com/temp/dist_bare.gif

The second case is a coil with a relatively large toroid installed. Note 
how the current changes, but only slightly.
http://www.classictesla.com/temp/dist_toroid.gif

The third case is a coil with a small toroid installed. Current doesn't 
go to zero, but is does change significantly.
http://www.classictesla.com/temp/dist_small_toroid.gif

In all cases, Ibase=(Vcap(pk) x sqrt(Les/Lp))/Sec Reactance. The 
calculated base current is only a near representation of the highest 
base current expected (it's not a value to size coil wire with). If you 
were to use a measured Ibase current, the current variation would not 
change, only the actual current scale.

The graphs make it easy to see how the top load geometry and size 
affects the distributed capacitance and how this changes the 
distribution of current along the length of the coil. If your wondering 
why base currents change with the same coil, cap, and transformer, 
realize the change in currents due to the distribution of capacitance 
affects the inductance from turn to turn. This is why I use the 
inductance at frequency (Les) within the base current equation.

Does it really make sense to change wire size along the coil? Most coils 
will fall between case 2 and 3, and the highest percentage of the coils 
will be nearer to case 2 where there is only a slight change. You gain 
nothing by winding the lower portion of the coil with larger wire and 
you actually lose out on inductance overall. If you must splice due to 
"what is on hand", then fine. But from a performance standpoint, I don't 
see a need to design for this.

Take care,
Bart

bartb wrote:
> I think Ed is correct. According to models, the larger the top load is 
> relative to the coil size, the "less" the current changes from top to 
> bottom. But if a bare coil, it is max to zero (or near). Also, the 
> linearity of the distributed current along the length of the coil is 
> dependent on the h/d (more linear for low h/d coils).
>
> For the normal h/d of 4 or 5, it's not quite linear, but curved even 
> with bare coils. Throw a big toroid on the coil and the current is far 
> less changing from bottom to top because the toroid is very well 
> adapted at forcing itself (it's C imposed on the inter C of the 
> windings) and changing the overall distributed capacitance of the coil 
> due to the toroid geometry. The current as a result becomes less 
> changing from bottom to top (but there is still a difference). So no, 
> current is not the same from end to end from a technical stand point. 
> And it can never be the same unless you throw turn to turn C out of 
> the equation.
>
> If you make a statement as "the top current is smaller than the bottom 
> current", then you would be correct. The problem is the person stating 
> this usually does not know "how much smaller". In rare cases, it can 
> be a lot. In most TC cases, it's not a big change. If I were to give a 
> "number", I would say the current on average coils is about 25% less 
> at the top than it is at the base. But, that's just truly rough 
> guesstimate of what I would call average coils and average tops to 
> coils size.
>
> Take care,
> Bart
>
> Ed Phillips wrote:
>> It's a long piece of wire with capacitance between turns and 
>> capacitance of each turn to the surroundings.  Current distribution 
>> depends on how much top load capacitance is on the coil.  With no top 
>> loading the current will be lower at the top; if there are no 
>> streamers the current should be essentially zero at the top.  
>> Distributed capacitance is the culprit.
>>
>> Ed
>>
> _______________________________________________
> Tesla mailing list
> Tesla at www.pupman.com
> http://www.pupman.com/mailman/listinfo/tesla
>